A surprising capacity of the adult mammalian eye to regrow (at least in vitro) has been discovered. Single retinal cells from the ciliary margin zone of the embryonic and adult mouse retina (as well as from the adult human retina) will proliferate in culture over 1 week to form spheres of 13,000 cells, that can both self-renew (several new spheres can be generated from the dissociation of each starting sphere) and produce progeny that can differentiate into all of the different neuronal and glial cell types in the retina. Our long term objective is to use these retinal stem cells to cure blindness. 3 specific aims will advance this objective. First, we suggest that retinal stem cells may be better purified by pigmentation, side population analysis and Pax6 expression, so that their differences from retinal progenitor cells and postmitotic retinal cells can be revealed. Second, what extrinsic and intrinsic factors control the activity of these mammalian retinal stem cells in the eye. Retinal stem cells from wild type mice and mice carrying genetic mutations that affect eye development will be cultured, in order to understand the factors that control retinal stem cell activity. We hypothesize that mutations that limit retinal progenitor proliferation result in the modulation of specific non-cell autonomous feedback signals (decreasing GDF11 and increasing FGF2) that activate stem cell proliferation. On the other hand, we hypothesize that Pax6 can work cell autonomously in the retinal stem cells themselves to regulate their activity. Third, adult mouse and human stem cells (after their isolation in culture) will be transplanted back into normal and damaged mouse eyes to test if the progeny of the transplanted retinal stem cells can replace missing retinal photoreceptors and functionally improve vision in mice with visual problems. This work is significant in that it sets the stage for using human retinal stem cells (either transplanted or endogenously activated) to repair the retinas of the blind. [unreadable] [unreadable] [unreadable]